Clinical outcomes from ART in predicted hyperresponders: in vitro maturation of oocytes versus conventional ovarian stimulation for IVF/ICSI

Abstract

Abstract STUDY QUESTION Do ongoing pregnancy rates (OPRs) differ in predicted hyperresponders undergoing ART after IVM of oocytes compared with conventional ovarian stimulation (OS) for IVF/ICSI? SUMMARY ANSWER One cycle of IVM is non-inferior to one cycle of OS in women with serum anti-Müllerian hormone (AMH) levels ≥10 ng/ml. WHAT IS KNOWN ALREADY Women with high antral follicle count and elevated serum AMH levels, indicating an increased functional ovarian reserve, are prone to hyperresponse during ART treatment. To avoid iatrogenic complications of OS, IVM has been proposed as a mild-approach alternative treatment in predicted hyperresponders, including women with polycystic ovary syndrome (PCOS) who are eligible for ART. To date, inferior pregnancy rates from IVM compared to OS have hampered the uptake of IVM by ART clinics. However, it is unclear whether the efficiency gap between IVM and OS may differ depending on the extent of AMH elevation. STUDY DESIGN, SIZE, DURATION This study is a retrospective cohort analysis of clinical and laboratory data from the first completed highly purified hMG (HP-hMG) primed, non-hCG-triggered IVM or OS (FSH or HP-hMG stimulation in a GnRH antagonist protocol) cycle with ICSI in predicted hyperresponders ≤36 years of age at a tertiary referral university hospital. A total of 1707 cycles were included between January 2016 and June 2022. PARTICIPANTS/MATERIALS, SETTING, METHODS Predicted hyperresponse was defined as a serum AMH level ≥3.25 ng/ml (Elecsys® AMH, Roche Diagnostics). The primary outcome was cumulative ongoing pregnancy rate assessed 10–11 weeks after embryo transfer (ET). The predefined non-inferiority limit was −10.0%. The analysis was adjusted for AMH strata. Time-to-pregnancy, defined as the number of ET cycles until ongoing pregnancy was achieved, was a secondary outcome. Statistical analysis was performed using a multivariable regression model controlling for potential confounders. MAIN RESULTS AND THE ROLE OF CHANCE Data from 463 IVM cycles were compared with those from 1244 OS cycles. Women in the IVM group more often had a diagnosis of Rotterdam PCOS (434/463, 93.7%) compared to those undergoing OS (522/1193, 43.8%), were significantly younger (29.5 years versus 30.5 years, P ≤ 0.001), had a higher BMI (25.7 kg/m2 versus 25.1 kg/m2, P ≤ 0.01) and higher AMH (11.6 ng/ml versus 5.3 ng/ml, P ≤ 0.001). Although IVM cycles yielded more cumulus–oocyte complexes (COCs) (24.5 versus 15.0 COC, P ≤ 0.001), both groups had similar numbers of mature oocytes (metaphase II (MII)) (11.9 MII versus 10.6 MII, P = 0.9). In the entire cohort, non-adjusted cumulative OPR from IVM was significantly lower (198/463, 42.8%) compared to OS (794/1244, 63.8%), P ≤ 0.001. When analysing OPR across different serum AMH strata, cumulative OPR in both groups converged with increasing serum AMH, and OPR from IVM was non-inferior compared to OS from serum AMH levels >10 ng/ml onwards (113/221, 51.1% (IVM); 29/48, 60.4% (OS)). The number of ETs needed to reach an ongoing pregnancy was comparable in both the IVM and the OS group (1.6 versus 1.5 ET’s, P = 0.44). Multivariable regression analysis adjusting for ART type, age, BMI, oocyte number, and PCOS phenotype showed that the number of COCs was the only parameter associated with OPR in predicted hyperresponders with a serum AMH >10 ng/ml. LIMITATIONS, REASONS FOR CAUTION These data should be interpreted with caution as the retrospective nature of the study holds the possibility of unmeasured confounding factors. WIDER IMPLICATIONS OF THE FINDINGS Among subfertile women who are eligible for ART, IVM, and OS resulted in comparable reproductive outcomes in a subset of women with a serum AMH ≥10 ng/ml. These findings should be corroborated by a randomised controlled trial (RCT) comparing both treatments in selected patients with elevated AMH. STUDY FUNDING/COMPETING INTEREST(S) There was no external funding for this study. P.D. has been consultant to Merck Healthcare KGaA (Darmstadt, Germany) from April 2021 till June 2023 and is a Merck employee (Medical Director, Global Medical Affairs Fertility) with Merck Healthcare KGAaA (Darmstadt, Germany) since July 2023. He declares honoraria for lecturing from Merck KGaA, MSD, Organon, and Ferring. The remaining authors declared no conflict of interest pertaining to this study. TRIAL REGISTRATION NUMBER N/A.